Not applicable.
(1) Field of the Invention
The present invention relates generally to systems and methods for a cross-correlator and, more specifically, for an adaptive cross-correlator operable for adaptively suppressing high signal to noise ratio (SNR) narrowband interference before the interference enters the cross-correlator process.
(2) Description of the Prior Art
A vessel""s sound signature contains both a continuous, broadband spectrum of sound, as well as discrete, narrowband tonals at specific frequencies along that spectrum that rise above the spectrum. The tonals may be caused by specific pieces of rotating machinery within the vessel. For instance, narrow band tonals may be produced by pumps, generators, and gears, whereas the continuous broadband spectrum is caused primarily by flow noise over the hull surface or by propeller cavitation. However, narrowband tonals or narrowband interference may also be produced by self noise related to a loud platform, such as the platform to which the sonar array is attached. Other narrow band tonals or narrowband interference may be produced by other vessels unrelated to the target vessel.
A vessel""s broadband signature may resemble background noise in that it contains a continuous spectrum of frequencies within which sound source levels at particular frequencies rise and fall in random fashion around a mean over time. By contrast, the narrowband component of a vessel""s signature may generate sound at several specific frequencies continuously. Thus, compared to the background noise generated at these specific frequencies, which will average out over time to x, the signal plus noise received at the tonal frequency will average out over time to x+y, with y being the source level of the signal.
At close range, a simple sonar will detect a vessel""s broadband signal simply by pointing the main beam of its array at the vessel. The sonar is measuring all the sound it receives in a given direction, including both signal and background noise, and as it points in the direction of the target, the signal increases. As the range of the opposing submarine is increased, the relative strength of this broadband signature compared to the broadband background noise declines until it is drowned out and the signal-to-noise ratio drops below the detection threshold. Therefore, self-noise is also an important issue for sonar effectiveness. This is true whether one is seeking broadband or narrowband detections.
In some cases, two important tonals in a submarine""s signature may be those modulated by the propeller at the rate which its blades turn, and those associated with particular items of rotating machinery. Blade rate tonals are usually slightly lower in frequency than machinery tonals, and both tonals are usually aspect dependent and speed dependent. Taken together, these tonals provide means to detect targets, classify them, and to track them over time.
High signal to noise ratio narrowband interference creates signal distortion and makes peak detection more difficult for received sonar signals. Removal or suppression of narrowband interference is not a new idea. The current methods for removal or suppression of narrowband interference from signals received by an array of sonar transducers utilize what is referred to as the smooth coherence transform (SCOT). More detailed information about SCOT methods can be found in references such as Coherence and Time Delay Estimation, by G. Clifford Carter, IEEE Press, Piscataway, N.J., 1992. SCOT works in the frequency domain and essentially applies a filter on the output of the frequency domain correlator implementation. However, SCOT tends to be computationally intensive. The number of floating point operations required per update are of order Mxc3x97N where N is 1024 or greater and M is the number of beam pairs. It would be desirable to provide a significant reduction in processing throughput as compared to the SCOT method. It has also been shown that in the presence of noise and wide band signals only, SCOT will have reduced performance over a cross correlator without SCOT processing (standard cross correlator). It is would therefore be desirable to provide an adaptive cross correlator that will perform no worse than the standard cross correlator in the presence of wide band signals and noise.
Various inventors have attempted to solve related problems as evidenced by the following patents:
U.S. Pat. No. 5,724,485, issued Mar. 3, 1998, to David Rainton, discloses an adaptive cross correlator apparatus with a first receiving section that receives a signal and outputs the received signal as a first signal, and a second receiving section that receives a further signal and outputs the received further signal as a second signal, wherein the second receiving section is provided at a position different from that of the first receiving section. A first filter filters the first signal with a first changeable transfer function and outputs a filtered first signal, and a second filter filters the second signal with a second changeable transfer function and outputs a filtered second signal. Further, a cross correlator calculates a cross correlation value by using a predetermined cross correlation function based on the filtered first and second signals, and then, an adaptive controller calculates a discriminant function value representing a misclassification measure of the first and second signals, based on the cross correlation value and a true delay between the first and second signals, and adaptively adjusts the respective first and second transfer functions of the first and second filters so that the calculated discriminant function value becomes a minimum.
U.S. Pat. No. 5,899,864, issued May 4, 1999, to Arenson et al., discloses the energy, power or amplitude of Doppler or time shift information signals that is compared to a threshold in order to select a large or small weighting factor for temporal persistence. In the event of a xe2x80x9cflashxe2x80x9d signal or strong arterial flow signal, a small weighting factor is chosen to reduce the extent of temporal persistence via feedback of the averaged value for the prior frames so that the effect of the xe2x80x9cflashxe2x80x9d or strong flow signal would quickly dissipate in the imaging of subsequent frames and good temporal resolution preserved for the current frames, while low energy flow signals would cause a large weighting factor to be selected to improve the signal-to-noise ratio of low energy signals. Similar effects can be achieved by clipping the signals to not exceed a certain threshold.
U.S. Pat. No. 6,130,643, issued Oct. 10, 2000, to Trippett et al., discloses an antenna nulling system for nulling a jamming signal having a multibeam antenna, a correlator, and antenna pattern calculator, a sequential updater and a beamformer. The multibeam antenna includes a plurality of antenna elements and is operable to receive the plurality of signals. The correlator is operable to receive at least one sample signal from one of the antenna elements and a composite signal from the plurality of antenna elements. The correlator determines a cross-correlation of the sample signal and the composite signal. The antenna pattern calculator calculates a difference in pattern magnitude of an adapted antenna pattern and a quiescent antenna pattern of the multibeam antenna. The sequential updater sequentially calculates a new weight for each of the antenna elements based upon an existing weight of each antenna element, the cross-correlation and the difference in pattern magnitude. The beamformer is in communication with the multibeam antenna and the sequential updater to combine a new weight for each of the antenna elements with the plurality of signals received from the multibeam antenna to null the jamming signal.
U.S. Pat. No. 5,978,473, issued Nov. 2, 1999, to Jim Agne Jerker Rasmusson, discloses a measure of a degree of convergence in an adaptive filter arrangement that is derived from the comparison of an amount of adaptation occurring in the adaptive filter arrangement, over a predetermined period of time, with a normalizing value accumulated for the same period. Supplemental signal processing may be invoked, modified or withdrawn based upon the degree of convergence indicated.
U.S. Pat. No. 5,901,343, issued May 4, 1999, to Julius Lange, discloses an intermediate frequency adaptive cross polarization interference canceller for processing an interfering cross polarization signal distorted by dispersion. The canceller has right and left inputs for respectively receiving right and left polarized IF input signals. A plurality of serially coupled complex multiplier and control stages respectively process the right and left polarized signals to provide controlled amounts of coupling between them to cancel cross polarization interference therebetween. A plurality of delay lines add predetermined time shifts to the right and left polarized signals between stages, and which forms a transversal filter having a predetermined number of taps. The canceller outputs right polarized IF output signal and a left polarized IF output signal having substantially no cross polarization interference therebetween. A preferred embodiment of the adaptive cross polarization interference canceller uses a compensating five-tap transversal filter disposed in the cancellation path. A simplified single tap adaptive cross polarization interference canceller may be used if there is no dispersion.
U.S. Pat. No. 5,852,567, issued Dec. 22, 1998, to Xia et al., discloses an iterative time frequency algorithm which filters noisy wide band/nonstationary signals by projecting the noisy signal into the TF domain, masking the TF response, computing the inverse TF transform to extract a filtered signal, and repeating these steps until the projection lies within the mask. As a result, the TF domain properties of the extracted signal are substantially equal to the desired TF domain properties. Furthermore, the iterative approach is computationally simple because it avoids inverting matrices. The TF transform and its inverse must be selected such that the iterative algorithm is guaranteed to converge. Candidate transform pairs can be tested on known data, and if the TF transforms converge to the desired TF properties, the candidate pair can be selected. Alternately, the candidate pairs can be tested against a sufficient convergence condition, and if they satisfy the condition within an acceptable tolerance, they can be selected with confidence. Furthermore, the sufficient convergence condition can be solved directly to provide the TF transform and its inverse.
U.S. Pat. No. 5,526,347, issued Jun. 11, 1996, to Chen et al., discloses a sign-based decorrelation detection and adaptive control arrangement which includes structure for detecting cross-correlation between a far-end signal and an echo residual following a balance filter. During the adaptive process, if the detected correlation value is below a certain threshold, indicating that the two signals are decorrelated, the adaptation of the balancing filter is stopped. At such a point, proper echo cancellation has been achieved. Conversely, when the detected correlation value exceeds a threshold, the adaptation is continued until the correlation value falls below the threshold again. In any event, such decorrelation controllers are able to detect signal decorrelation and to control adaptation even in the presence of a double-talker condition.
U.S. Pat. No. 5,416,532, issued May 16, 1995, to Jung W. Ko, discloses horizontal and vertical peaking signals that are separated from a video signal by combining variously delayed responses to the video signal. A cross-fader combines the separated horizontal and vertical peaking signals in proportions determined by a cross-fader control signal. A correlator responds to ones of the variously delayed responses to the video signal for generating an output signal representative of the relative degrees of vertical and horizontal correlation in the video signal. The correlator output signal addresses a read-only memory that supplies the cross-fader control signal. The adaptively generated peaking signal is suitable for adjustably peaking a luminance component extracted from the video signal, where that video signal is a composite signal also including a chrominance component. The extraction of the luminance component is preferably done on an adaptive basis, generating horizontal and vertical comb filter responses by suitably combining the variously delayed responses to the video signal and, with a further cross-fader, combining the horizontal and vertical comb filter responses in proportions determined by the cross-fader control signal.
U.S. Pat. No. 5,016,261, issued May 14, 1991, to Amoroso et al., discloses a method and apparatus for improving the anti-jam performance of a processing circuit to increase conversion gain and reduce small signal suppression resulting from processing a phase modulated input signal accompanied by jamming interference. Input signals are each segregated into signal chips by a matched filter, and by an adaptive threshold circuit in accordance with a predetermined relative threshold. The threshold is set to repeatedly distinguish a predetermined number of signal chips having greater signal amplitude. The absolute amplitude threshold level may, therefore, vary in accordance with the particular signal chips forming each input signal segment. A phase quantizer operates to extract phase information from the signal chips. A phase correlator operates to apply a first weighting gain factor to signal chips equal to or exceeding the threshold, and a second weighting gain factor to the remaining chips, the first weighting gain factor being greater than the second weighting gain factor.
U.S. Pat. No. 4,270,179, issued May 26, 1981, to Sifford et al., discloses a complex ternary correlator and method for adaptive gradient computation in an adaptive equalizer which includes four ternary operation circuits, four ternary multiplier circuits for obtaining the cross products of the ternary operation outputs, a subtractor circuit for developing a signal commensurate with the difference between two of the ternary multiplier outputs, an adder circuit for developing a signal commensurate with the sum of the remaining two ternary multiplier outputs and two identical integrating circuits for obtaining the real and imaginary adaptive tap coefficient update increments in an adaptive equalizer.
U.S. Pat. No. 3,882,498, issued May 6, 1975, to August L. McGuffin, discloses an AMTI adaptive array in which each array antenna element is connected to an element circuit which multiplies the contribution of each antenna element to the total return m by a weight. The element circuits and further signal processing circuitry comprise the array processor. The element circuits include well-known cross correlator control loops. A signal whose pulse repetition interval to pulse repetition interval Doppler phase shift is 180xc2x0 out of phase with clutter returns is supplied to a control loop. Thus even with clutter at or near the look angle, main lobe gain is maintained. Consequently, a xe2x80x9ctwo pulsexe2x80x9d MTI circuit utilizing a single delay line in each element circuit may be utilized rather than a two delay line element circuit which would normally be required to provide a signal to noise ratio value indicative of main lobe gain which would be required for compatibility with further MTI processing circuitry.
The above patents do not provide an adaptive cross-correlator that requires no feedback from the correlator output or any prior knowledge of the expected process output. Consequently, there remains a long felt but unsolved need for an adaptive filter to suppress high signal to noise ratio narrowband interference while preserving the broadband energy before the signal enters the cross-correlation process. Those skilled in the art will appreciate the present invention that addresses the above and other problems.
Accordingly, it is an objective of the present invention to provide an improved sonar signal processing system and method.
Another objective is to provide a system and method as aforesaid which is operable to estimate narrowband interference in each of a plurality of beam signals and then subtract the estimated narrowband interference from each respective beam signal to thereby provide a plurality of filtered beam signals containing suppressed narrowband interference for application to a cross correlation processor.
A further objective is to provide a system and method as aforesaid whereby the narrowband interference is suppressed without the need of feedback from the cross correlation processor to which the filtered beam signals are applied.
These and other objectives, features, and advantages of the present invention will become apparent from the drawings, the descriptions given herein, and the appended claims. However, it will be understood that above listed objectives and advantages of the invention are intended only as an aid in understanding aspects of the invention, are not intended to limit the invention in any way, and do not form a comprehensive list of objectives, features, and advantages.
In accordance with the present invention, a system is provided for a sonar signal detection system that may comprise one or more elements such as, for instance, a plurality of sonar sensors, and/or a beamformer for receiving acoustic signals from the plurality of sonar sensors. The beamformer preferably produces a plurality of beam outputs. The beam outputs, x(n), have narrowband signal components ŜNB(n) and wideband signal components. A plurality of adaptive filters receive the plurality of beam outputs. Each of the plurality of adaptive filters has a delay component providing a delay, D, operable for decoupling the wideband signal component from the original wideband signal component. An adaptive notch filter is utilized to allow passage of the narrowband signal component and to suppress the wideband component. The delayed, notch filtered signal is then subtracted from the original signal and provided to a cross correlation processor. The plurality of adaptive filters are preferably operable without feedback from the cross correlation processor.
The system may further comprise a finite impulse response (FIR) filter for receiving the delayed narrowband signal, the delayed wideband signal, and the delayed ambient noise signal. The FIR filter output of each the FIR filters may be of the form:                                                                         S                ^                            NB                        ⁡                          (              n              )                                =                                    ∑                              k                =                0                                            P                -                1                                      ⁢                                          h                ⁡                                  (                  k                  )                                            *                              x                ⁡                                  (                                      n                    -                    k                    -                    D                                    )                                                                    ,                            (        1        )            
wherein h(k) comprises a plurality of filter coefficients.
In a preferred embodiment, the plurality of filter coefficients are adjusted recursively using a least mean squares (LMS) method of the form:
hn(k)=hnxe2x88x921(k)+xcex94*(x(n)+ŜNB(n))*x(nxe2x88x92kxe2x88x92D),xe2x80x83xe2x80x83(2)
where
0 less than xcex94 less than 1/(N*10*"sgr"x2)xe2x80x83xe2x80x83(3)
where "sgr"x2 is an estimate of power in input signal x(n).
In operation, a method for processing sonar signals may comprise one or more method steps such as, for instance, utilizing a beamformer to produce a plurality of beam outputs, and splitting each of the plurality of beam outputs into a first path and a second path. In the first path, the method preferably comprises applying the plurality of beam outputs to a plurality of signal combiners, and in the second path, the method comprises delaying the plurality of beam outputs with respect to time to produce a plurality of delayed beam outputs. Other steps may comprise applying the plurality of delayed beam outputs to a plurality of adaptive notch filters to produce a plurality of adaptive notch filter outputs, applying the plurality of adaptive notch filter outputs to the plurality of signal combiners, and producing a plurality of signal combiner outputs for application to a cross correlation processor.
In a preferred embodiment, the method may comprise producing the plurality of signal combiner outputs without feedback from the cross correlation processor. The method may further comprise displaying an output of the cross correlation processor.
In one embodiment, the method may comprise providing that the plurality of beam outputs comprise a narrowband signal, a wideband signal, and ambient noise signal. The method comprises suppressing the narrowband signal while passing the wideband signal and the ambient noise signal to the cross correlation processor.
In other words, the system provides for a plurality of adaptive filters for receiving the plurality of beam outputs, wherein each of the plurality of adaptive filters may comprise a finite impulse response filter, a plurality of delay components for the plurality of adaptive filters, and a plurality of signal combiners for the plurality of adaptive filters.